Forum Discussion
Text-to-Speech Pronunciation
Does anyone have any advice on how to get the text-to-speech AI to pronounce WAN to it rhymes with ban? I'm using the voice "Matilida," if that matters. Here's what I've tried so far:
Wahn, Wann, Whann, Whanne, Wanne, Wannn
Please help!
Thanks, all. I wasn't able to get it there with either of these suggestions, but I spelled it Huann, which resulted in a pronunciation of who-wan, and then I just trimmed off the first part. Mission accomplished, even if it took a few detours!
7 Replies
I love Matilda, but she's stubborn at times. Try w'An.
Have you tried adding SSML tags to adjust the pronounciation?
I have had both good and less than optimal results.
https://access.articulate.com/support/article/Storyline-360-SSML-SupportThanks, all. I wasn't able to get it there with either of these suggestions, but I spelled it Huann, which resulted in a pronunciation of who-wan, and then I just trimmed off the first part. Mission accomplished, even if it took a few detours!
PLEASE HELP!
Matilda is my Go-To but I am having a heck of a time trying to get her to say "executor" as in the executor of a will. I have tried SSML
<speak>
<phoneme alphabet="ipa" ph="ɛɡˈzɛkjʊtɚ">executor</phoneme>
</speak> (This is ignored altogether and silence is inserted.)and
<speak>
<emphasis level="strong">exECutor</emphasis>
</speak> (here she says egzecuTOR with an asian accent.)as well as numerous phonetical spellings but she just ignores them.
interestingly, I tried Aria, Emily, Elli, Jessica, Sarah, Rachel and several others, all pronounce it properly. just Matilda who I am using on a very big project and the client loves her.
HOW DO I FIX THIS???
Hi RobSkeet,
Thanks for reaching out about this!
What about spelling it as exectcuter? Here's my test. Others in the community may have some different ideas!
OK, very weird. I tried it in the full text and got exec cute tutor
but when I tried it on its own using your "exectcuter of a will" it worked fine. so I exported that to a file then edited the word back into the OG. that worked as long as I don't have to edit the text down the road. I was then curious and just did "executor of a will" on its own and it also worked. so it must be how Matilda is seeing the word in the context of the paragraph. So, AI assistant voices don't incorporate SSML?
Hi RobSkeet,
I'm happy to chime in!
You're absolutely right. AI Assistant has limited support for Speech Synthesis Markup Language (SSML) because AI-generated voices are designed to understand the relationship between words and adjust delivery accordingly. In this case, Matilda reads the phrase "executor of a will" correctly by recognizing the context between the words.
Let us know if you have any more questions or if there's anything else we can help with.
